Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost

Qunxiong Zhu; Xinwei Wang; Ning Zhang; Yuan Xu; Yanlin He

doi:10.1109/DDCLS55054.2022.9858431

Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost

Qunxiong Zhu, Xinwei Wang, Ning Zhang, Yuan Xu, Yanlin He

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Citations (Scopus)

Abstract

The data collected in modern process industry have imbalanced, high-dimensional, and non-linear features, which bring great challenges to chemical process fault diagnosis. Facing these features of data, we present a new fault diagnosis method based on Borderline Synthetic Minority Over-Sampling Technique (BorSMOTE) integrated with Neighborhood Preserving Embedding (NPE) and CatBoost named BSNC. In the proposed BSNC, BorSMOTE is an improved oversampling method based on SMOTE, which improves the class distribution of samples by using only minority class sample on the boundary to synthesize some new samples; NPE is used for dimensionality reduction (DR) to extract critical features associated with faults; finally, CatBoost is used as a classifier to identify the fault types. In order to verify the feasibility of the proposed BSNC methodology, the Tennessee Eastman process (TE) with different types of fault data is chosen for simulation experiment validation. The simulation results show that the BSNC methodology in this paper has considerable performance compared with the data in the imbalanced state and the related DR methodologies.

Original language	English
Title of host publication	Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022
Editors	Mingxuan Sun, Zengqiang Chen
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	612-617
Number of pages	6
ISBN (Electronic)	9781665496759
DOIs	https://doi.org/10.1109/DDCLS55054.2022.9858431
Publication status	Published - 2022
Externally published	Yes
Event	11th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2022 - Emeishan, China Duration: 3 Aug 2022 → 5 Aug 2022

Publication series

Name	Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022

Conference

Conference	11th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2022
Country/Territory	China
City	Emeishan
Period	3/08/22 → 5/08/22

Keywords

Borderline SMOTE
CatBoost
Imbalanced Fault Diagnosis
Neighborhood Preserving Embedding
Tennessee Eastman

Access to Document

10.1109/DDCLS55054.2022.9858431

Cite this

Zhu, Q., Wang, X., Zhang, N., Xu, Y., & He, Y. (2022). Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost. In M. Sun, & Z. Chen (Eds.), Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022 (pp. 612-617). (Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/DDCLS55054.2022.9858431

Zhu, Qunxiong ; Wang, Xinwei ; Zhang, Ning et al. / Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost. Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022. editor / Mingxuan Sun ; Zengqiang Chen. Institute of Electrical and Electronics Engineers Inc., 2022. pp. 612-617 (Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022).

@inproceedings{ae08e654bcac479596dab0c228f204d7,

title = "Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost",

abstract = "The data collected in modern process industry have imbalanced, high-dimensional, and non-linear features, which bring great challenges to chemical process fault diagnosis. Facing these features of data, we present a new fault diagnosis method based on Borderline Synthetic Minority Over-Sampling Technique (BorSMOTE) integrated with Neighborhood Preserving Embedding (NPE) and CatBoost named BSNC. In the proposed BSNC, BorSMOTE is an improved oversampling method based on SMOTE, which improves the class distribution of samples by using only minority class sample on the boundary to synthesize some new samples; NPE is used for dimensionality reduction (DR) to extract critical features associated with faults; finally, CatBoost is used as a classifier to identify the fault types. In order to verify the feasibility of the proposed BSNC methodology, the Tennessee Eastman process (TE) with different types of fault data is chosen for simulation experiment validation. The simulation results show that the BSNC methodology in this paper has considerable performance compared with the data in the imbalanced state and the related DR methodologies.",

keywords = "Borderline SMOTE, CatBoost, Imbalanced Fault Diagnosis, Neighborhood Preserving Embedding, Tennessee Eastman",

author = "Qunxiong Zhu and Xinwei Wang and Ning Zhang and Yuan Xu and Yanlin He",

note = "Publisher Copyright: {\textcopyright} 2022 IEEE.; 11th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2022 ; Conference date: 03-08-2022 Through 05-08-2022",

year = "2022",

doi = "10.1109/DDCLS55054.2022.9858431",

language = "English",

series = "Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "612--617",

editor = "Mingxuan Sun and Zengqiang Chen",

booktitle = "Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022",

address = "United States",

}

Zhu, Q, Wang, X, Zhang, N, Xu, Y & He, Y 2022, Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost. in M Sun & Z Chen (eds), Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022. Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022, Institute of Electrical and Electronics Engineers Inc., pp. 612-617, 11th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2022, Emeishan, China, 3/08/22. https://doi.org/10.1109/DDCLS55054.2022.9858431

Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost. / Zhu, Qunxiong; Wang, Xinwei; Zhang, Ning et al.
Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022. ed. / Mingxuan Sun; Zengqiang Chen. Institute of Electrical and Electronics Engineers Inc., 2022. p. 612-617 (Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost

AU - Zhu, Qunxiong

AU - Wang, Xinwei

AU - Zhang, Ning

AU - Xu, Yuan

AU - He, Yanlin

PY - 2022

Y1 - 2022

N2 - The data collected in modern process industry have imbalanced, high-dimensional, and non-linear features, which bring great challenges to chemical process fault diagnosis. Facing these features of data, we present a new fault diagnosis method based on Borderline Synthetic Minority Over-Sampling Technique (BorSMOTE) integrated with Neighborhood Preserving Embedding (NPE) and CatBoost named BSNC. In the proposed BSNC, BorSMOTE is an improved oversampling method based on SMOTE, which improves the class distribution of samples by using only minority class sample on the boundary to synthesize some new samples; NPE is used for dimensionality reduction (DR) to extract critical features associated with faults; finally, CatBoost is used as a classifier to identify the fault types. In order to verify the feasibility of the proposed BSNC methodology, the Tennessee Eastman process (TE) with different types of fault data is chosen for simulation experiment validation. The simulation results show that the BSNC methodology in this paper has considerable performance compared with the data in the imbalanced state and the related DR methodologies.

AB - The data collected in modern process industry have imbalanced, high-dimensional, and non-linear features, which bring great challenges to chemical process fault diagnosis. Facing these features of data, we present a new fault diagnosis method based on Borderline Synthetic Minority Over-Sampling Technique (BorSMOTE) integrated with Neighborhood Preserving Embedding (NPE) and CatBoost named BSNC. In the proposed BSNC, BorSMOTE is an improved oversampling method based on SMOTE, which improves the class distribution of samples by using only minority class sample on the boundary to synthesize some new samples; NPE is used for dimensionality reduction (DR) to extract critical features associated with faults; finally, CatBoost is used as a classifier to identify the fault types. In order to verify the feasibility of the proposed BSNC methodology, the Tennessee Eastman process (TE) with different types of fault data is chosen for simulation experiment validation. The simulation results show that the BSNC methodology in this paper has considerable performance compared with the data in the imbalanced state and the related DR methodologies.

KW - Borderline SMOTE

KW - CatBoost

KW - Imbalanced Fault Diagnosis

KW - Neighborhood Preserving Embedding

KW - Tennessee Eastman

UR - http://www.scopus.com/inward/record.url?scp=85137787921&partnerID=8YFLogxK

U2 - 10.1109/DDCLS55054.2022.9858431

DO - 10.1109/DDCLS55054.2022.9858431

M3 - Conference contribution

AN - SCOPUS:85137787921

T3 - Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022

SP - 612

EP - 617

BT - Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022

A2 - Sun, Mingxuan

A2 - Chen, Zengqiang

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 11th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2022

Y2 - 3 August 2022 through 5 August 2022

ER -

Zhu Q, Wang X, Zhang N, Xu Y, He Y. Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost. In Sun M, Chen Z, editors, Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022. Institute of Electrical and Electronics Engineers Inc. 2022. p. 612-617. (Proceedings of 2022 IEEE 11th Data Driven Control and Learning Systems Conference, DDCLS 2022). doi: 10.1109/DDCLS55054.2022.9858431

Industrial Imbalanced Fault Diagnosis Method Based on Borderline SMOTE Integrated with NPE and CatBoost

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this